Boost Regulator [Off Topic, i know]

[hmmpic]

Hi

@Sorry for my Off topic, but all the clever mind is in here...

Working on a system where my software is running nice:grin:

Problem is my hardware.
System is a Data TX/RX system there are working on a 3,6 V Li-SOCl2 cell.

Main problem is cold -30 degree where the Li-SOCl2 is not delivering the voltage there are needed, it was expected from the data sheet.

When CPU and system are in sleep mode it only drain 33uA
When in TX mode for about 300ms it drain 100mA and in -30 deg. voltage drop. When using it from -5 and up to normal temp. no problem.

Now i thinking about using BOOST like MCP16251 from Microchip.
https://www.mouser.dk/ProductDetail/579-MCP16251T-I-CH


**Question is,
if input are dropping from the about 3,3 to ex 2,9 and the BOOST is set to deliver 3,3 on the output, will is then work?
I mean will the BOOST reg compensate the output drop?

[Ttelmah]

Yes, _but_. There is a problem here. This is a _boost_
regulator. It can only produce an output above it's input voltage. Problem
is that if your battery starts to produce it's 'warm' output of perhaps
3.6v, the regulator output stage will be reverse biased....

You need to be using a buck-boost regulator or a Cuk regulator.

Now you also need to consider that in order to give the extra output voltage, more current will be drawn at the low input voltage, and this is made worse by the inherent efficiency loss. So for 3.3v out off a battery voltage of 2.6v, to deliver 10mA, the regulator will draw perhaps 14mA... Result battery will discharge quicker....

[hmmpic]

Thanks for your feedback.
In the data sheet there are a note 2 on page 3, "Output Voltage Adjust Range:" (Note 2: For VIN > VOUT, VOUT will not remain in regulation.)
-For me it sound like the unit is just passing the input direct to the output.
But thanks for pointing it out:-)

[Ttelmah]

The output transistor becomes reverse biased. At enough voltage it'll start to
act as a diode, and the output the will go up without regulation, but with input
slightly 'above' the required voltage, the output will actually drop until the
diode voltage is reached, so regulation is then lost. It is also endangering the
part. It is not designed to do this....
It really is a "don't use" mode.

[hmmpic]

Thanks.
I will redesign the power interface, and use a 1,5V "Energizer Ultimate Lithium" it look nice down to -40 degree.

[Ttelmah]

All batteries lose both voltage and capacity at lower temperatures.
The Lithium Disulphide chemistry has one of the smallest effects but still
quite major.
The output voltage falls by typically about 0.2v at -40, and the capacity is
only about 40% of the room temperature value. Remember the peak
voltage from this chemistry can be as much as 1.8v. The nominal figure
can be quite misleading. You have to design your circuitry to be able to
use down to 0.8v to use the full capacity of these cells.
Choose your pack to have at least twice the capacity you want when at
normal temperatures. Choose a voltage that will always be above what you
want even at the lowest temperature. Then you can use a simple buck
regulator. Alternatively choose a voltage that will always be below what
is required and go with the boost regulator. Remember though the much
higher current that will then be needed.

[temtronic]

Microchip appnote AN1311 will be of interest. A single cell boost converter design. There probably are a few ready made 'modules' available, though not sure if they'll survive the -30*C spec you nedd, brrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr !
You should be able to get detailed specs for the ultimates, if not, give me a shout, I have a great contact at Energizer.
You could parallel 2 of the cells. Out of a fresh pack of 4 they were 1.795, 1.794, 1.795, 1.802.

Jay